Industrial automation systems are used to monitor, control and regulate technical processes, such as in the field of production, process and building automation, and make it possible to operate control devices, sensors, machines and industrial installations, which should be performed as autonomously and independently of human intervention as possible. Due to an ever increasing importance of information technology for automation systems, which comprise numerous networked computer units, methods for reliably providing functions distributed over an automation system are becoming increasingly important for the purpose of providing monitoring, control and regulating functions.
In the case of distributed services, whose service components are provided by a plurality of different computer units for implementing functions of the automation system, high demands are imposed on a communication infrastructure of an automation system. On the one hand, it is usually necessary to satisfy demands for real-time communication which is characterized by a comparatively large number of relatively short messages, and, on the other hand, the intention is to be able to use different communication protocols for distributed services as transparently as possible. An inefficient communication infrastructure may jeopardize error-free provision of a service and may have a disadvantageous effect on system stability and availability.
In addition, highly different client demands are often imposed on availability and quality of service, as well as data filtering and information pre-compression, when accessing data and measured values associated with services within an industrial automation system. Furthermore, there are demands for existing services to be able to be extended by optional or additional service components without having to substantially change a data and function model on which implementation of an existing service is based. These demands were previously satisfied by essentially separately implementing functions for ensuring a defined service availability or data filters for all services on the service side within an industrial automation system. This is extremely disadvantageous with regard to efficiently providing and maintaining the functions. Separate individual solutions are also problematic with regard to a consistent system behavior in the case of distributed services.
Previous approaches provide a specific communication adapter which is assigned to an individual service control unit and is adapted to the individual service control unit. Therefore, far-reaching adaptation work on the communication adapter is often required in the event of changes to components of a service managed by the service control unit.